You are on page 1of 9

Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010

The ‘Resource Curse’: Theory and Evidence (ARI)
Jonathan Di John*
Theme: Mineral and fuel abundance does not determine either the political or economic trajectory of less developed countries.

Summary: This paper undertakes a critical survey of the ‘resource curse’ –the idea that mineral and fuel abundance generates poor economic performance in less developed countries–. It examines the proposition that mineral and fuel abundance generates growth-restricting forms of state intervention and extraordinarily large degrees of rentseeking and corruption, which are generally argued to be negative in terms of the economic growth outcomes they generate. The analysis surveys the Dutch Disease, rentier state, and rent-seeking versions of the resource curse and finds they have significant shortcomings in terms of both theory and evidence. It also discusses particular growth strategies that have been effective in producing long-run economic growth in mineral- and fuel-abundant developing countries.

Analysis: Introduction One of the more influential ideas in recent development discourse and policy is the socalled ‘resource curse’. The big idea behind the ‘resource curse’ is that mineral and fuel abundance in less developed countries (LDCs) tends to generate negative developmental outcomes, including poor economic performance, growth collapses, high levels of corruption, ineffective governance and greater political violence. Natural resources, for most poor countries, are deemed to be more of a ‘curse’ than a ‘blessing’. In terms of intellectual history, this negative view of mineral and fuel abundance goes against much of the earlier thinking on the subject. Many analysts suggested a historically positive association of natural resource abundance and industrial growth in many now advanced countries. For instance, the ‘staple thesis’ demonstrated that growth in backward areas commonly began through the initial stimuli that primary product exports brought in terms of attracting capital and labour and inducing a more diversified production structure (Innis 1930; Watkins 1963). Also, natural resource rents, to the extent they are appropriated by state governments, can relax common resource constraints to growth –namely the savings, foreign exchange and fiscal constraints (Gelb & Associates, 1988, p. 17-18)–.1

*
1

School of Oriental and African Studies (SOAS), University of London.

Findlay & Lundahl (1999) find a generally growth-enhancing role of natural resource-rich countries in the period 1870-1914.

1

Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 This paper provides a critical survey of the theory and evidence with respect to some of the main versions of the ‘resource curse’. 2004. One of the potential dangers of oil booms. however. when a late-developing country faces a technological gap.2 It suggests some of the reasons why much of the evidence is inconclusive. With these assumptions. The mechanism through which this change takes place follows directly from the model’s assumptions of full employment equilibrium and static technology. can play an 2 The issue of whether mineral abundance generates greater political violence has been treated elsewhere (see Ross. 1967). it is a ‘blueprint’). which means that additional foreign exchange is not of particular relevance from the point of view of economic growth. an appreciation of the exchange rate leads to a decline in the competitiveness. 2 . 3 Sachs & Warner (1995) find. using the Sachs & Warner data. It focuses on the extent to which mineral abundance affects economic performance. opens up an important role for policy in affecting the growth outcomes of oil booms. That is. In the short-run.3 It also provides some policy implications that emerge from the discussion.4 A second mechanism through which manufacturing can become less competitive in this model is through the increase in manufacturing wage rates that result from increases in aggregate demand for labour that the oil booms can generate. in the period 1971-89. technology is assumed to be given (ie. Lederman & Maloney (2007). 4 Dutch Disease models are summarised in Neary & van Wijnbergen (1986). The logic of the simple Dutch Disease theories can be described as follows. since non-traded goods cannot be imported easily (or only at prohibitive costs). which can. leading to the crowding out of non-oil tradeables by non-tradeables. for example. that mineral exporters. in the absence of compensating policies. However. a relative contraction of the traded-goods sector is inevitable. when productivity levels are fixed. Di John. Thus. unit labour costs in manufacturing rise. find that there is not robust evidence to suggest that resource abundance negatively affects growth. Variants of the Resource Curse Argument (1) The Dutch Disease Model: the economic concept of Dutch Disease refers to the potential negative effects natural-resource windfalls and accompanying appreciations of exchange rates can have for the rest of the economy. In an economy in full employment equilibrium. a permanent increase in the inflow of external funds results in a change in relative prices in favour of non-traded goods (services and construction) and against non-oil traded goods (manufacturing and agriculture). However. on average. The potential dynamism that manufacturing can generate. of the traded-goods sector. However. The association of ‘de-industrialisation’ as a ‘disease’ stems from the unique growthenhancing characteristics the manufacturing sector can potentially embody (Kaldor. 2007). lead to a loss in manufacturing competitiveness. grew more slowly than the average growth of non-mineral exporters. otherwise the resources needed to enhance the growth of the nontraded sector would not be available. additional export revenues. is that exchange-rate appreciation renders the non-oil-tradeable sectors such as manufacturing less competitive and thus can generate deindustrialisation. the model predicts that de-industrialisation is the inevitable structural change that occurs as a result of oil booms. the external funds (from an oil boom) can be translated into real domestic expenditure only if the flow of imports increases. In the simple Dutch Disease model. if channelled by an appropriate industrial policy. and hence production and employment.

(2) Models of the Rentier State: rentier-state models move beyond economic models of the resource curse. they attempt to explain why state decision-makers in natural resource-rich economies create and maintain growthrestricting policies. the government could promote industry by channelling resources to toward that sector through protection. 3 . p. 8 For example. if the industrial strategy promotes ‘learning’. This in turn reduces political bargaining between state and interest groups. Di John (2009. or institutional evolution of a society–. Neary & van Wijnbergen note: ‘In so far as one general conclusion can be drawn [from our collection of empirical studies] it is that a country’s economic performance following a resource boom depends to a considerable extent on the policies followed by its government… [E]ven small economies have considerable influence over their own economic performance’. structural change against non-oil tradeables. the outcomes resource booms depend on state policy responses.5 Evidence from Venezuela. rather. such as Dutch Disease models. 7 For example. more importantly. the need to keep bribes secret reduces the security of property rights.6 What the Dutch Disease literature does not address is why growth-enhancing policies are chosen in some contexts and not others and. financial incentives and investments in infrastructure. Supporters of the rentier-state model suggest that reducing a state’s ‘unearned income’ from mineral rents will enhance the prospects of peace. For instance.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 important part since the additional foreign exchange can accelerate the process of importing advanced technology and the machines that embody them. There are several important propositions that are developed within this framework. increases in rent-seeking and corruption generate lower growth. the absence of incentives to tax internally weakens the administrative reach of the state. 35-76). oil and mineral abundance is assumed to generate growthrestricting state intervention and extraordinarily large degrees of rent-seeking. during the boom. paternalistic and even predatory. is not inevitable. In particular. the existence of a higher level of mineral rents increases rent-seeking and corruption relative to economies with lower mineral abundance. see Mahdavy (1970). Policy recommendations include advocating greater transparency in the payments made by multinationals in extractive 5 6 Neary & van Wijnbergen (1986. As a result. Fourth. where these rent-seeking contests are assumed to be uniformly negative in terms of the developmental outcomes they generate. such as manufacturing. which makes governance more arbitrary.7 These models are part of a growing trend of reviving the ‘staples thesis’ –the notion that natural factors endowments or technology shape the relations of production. This is in part due to the fact that with corrupt transactions. This can serve to modernise the manufacturing capital stock which in turn can improve productivity. additional revenues can theoretically accelerate the growth process. 10-11). Karl (1997) and Auty (2007). Additionally. capacity and legitimacy to intervene in the economy. why some leaders do not correct ineffective policies. see Engerman & Sokoloff (1997). by attempting to endogenise policymaking and institutional formation. oil rents provide a sufficient fiscal base of the state and thus reduce the necessity of the state to tax citizens. which lowers investment in long-gestating projects. Third. for instance. First.8 In the rentier-state model. suggests that policy responses (such as industrial policy and exchange-rate management) determine how oil booms affect the growth prospects of the economy. Secondly. p. subsidies. which results in lower levels of state authority.

because most revenues originate in the central government. the level of state discretion in allocating resources and regulating the economy tend to be higher than in most non-oil economies. the rentier-state theory cannot explain the long-run variation and change in growth of mineral abundant economies (eg. India. Third. Secondly. Critiques of Rent-seeking Theory The extent to which mineral economies generate both higher rent-seeking costs and less developmental rent-seeking outcomes is ultimately an empirical issue. In terms of the relationship between corruption and growth. Moreover. Tanzania and Malawi) either. First. the variation and change in economic growth in non-mineral rich economies is not well explained (eg. see Khan & Jomo (2000). Mozambique. The fact that aid-dependent economies may be pursuing more liberal economic policies demonstrates that policy matters more than levels of rents in the economy. China. the evidence in the Table suggests that corruption rates are indeterminate with respect to long-run growth. Uganda.9 In oil economies. Malaysia.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 industries to host governments in poor countries (Center for Global Development. there is also little support for the rent-seeking variant of the ‘resource curse’. Botswana. rather than through bargains between private economic agents. Venezuela and Nigeria). There are several pieces of evidence to suggest that large inflows of resources (whether through oil or aid) lead to a worsening in economic performance. 4 . 2004. The basic idea behind these models is that there are substantial costs to the workings of an economy when the allocation of resources is channelled primarily through state leaders. Tanzania and Ghana). who have discretionary authority. recent growth accelerations in aid-dependent economies is not well explained (eg. In the rentier-state model. the predominant view is that oil economies are subject to a higher level of rent-seeking and corruption in comparison with non-mineral abundant economies. or avoiding extractive industries altogether and concentrating efforts in order to diversify mineral-dominant economies towards agriculture and manufacturing (Ross. 9 For a critical survey of rent-seeking theory. Table 1 suggests that mineral-abundance economies do not appear to be more corrupt than non-mineral abundant economies. Modern theories of rent-seeking and corruption form a substantial part of the intellectual foundation of the rentier-state model. p. although there is considerable debate as to whether liberal economic policies are best for less-developed countries. 56-7). 2001). Let us consider these issues in more detail.

Non-Mineral-Abundant Developing Countries (2) Developing Countries (2) (13 observations) (19 observations) Median GDP Growth 4. Subjective Corruption indices from Transparency International. analysing the processes through which rights are assigned. such as ‘a predator’ or ‘rent-seeking maximiser’. None of this evidence provides much support for the rentier-state and rent-seeking models.7 (-0. non-mineral abundant is defined as those economies where mineral/fuel exports in total exports is less than 35% in 1980. 5 . when the real problem of common pool resources is.3%).2 (1.6 .8) Median GDP Growth Rate 1990-2000 (Range) Median Corruption Index 1996 (Range) 1. Why a particular coalition in power will not use oil revenues to diversify production is not addressed.6 Rate 1965-90 (2.5 .0 .4%) did outpace the mineral-abundant economies (4. the whole problematic of how to manage ‘common pool resources’ is neglected. Growth and Corruption in Mineral-Abundant and Non-Mineral Abundant Developing Countries. in fact.0) 2.0) (1) A corruption index of 10 indicates minimum corruption. Sources: World Bank. (2) Mineral-abundant is defined as those economies where mineral/fuel exports in total exports in 1980 is equal or greater to 35%. Secondly. leaders are assumed to have predatory as opposed to developmental aims.6 (0. Mineral-Abundant Developing Countries (13 observations) 4. but a set of social relations.9. maintained and changed (Ostrom. In the period 1965-90 the median annual average growth of the non-mineral abundant developing economies (5. enforced.8) 3.7 . an index of 0 indicates maximum corruption. The neglect of the political processes through which a leader appropriates power limits our understanding of the motivations of state leaders. World Development Indicators. First. By assigning ‘rights’ to leaders.4) (1. Mineral-Abundant 2. In other words.3 5.2 . The state is not a thing. it is assumed that there are no collective actors within the society that can impose some domestic conditionality on how those who occupy the state exercise their power.6.5. How rulers appropriate and maintain power is not analysed.7.8.10.12. 1990). they are assigned the ‘property rights’ over resources. rulers are assumed to ‘own’ the natural resources.3) 3. Critiques of Rentier-State Theory There are several assumptions of the rentier-state argument as developed by Terry Karl that drive the results.6 . Non-Mineral-Abundant Developing Countries (19 observations) 3. In the period 19902000 the mineral-dominant economies grew slightly faster and were slightly less corrupt than the non-mineral dominant economies. the median corruption rate of the non-mineral dominant economies was slightly higher than the mineral-dominant economies.7 .5 . 1965-2000 1965-1990 1.6.5) (Range) Median Corruption Index 1980-85 (1) (Range) 1990-2000 3.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 Table 1.5) 3. However.0 (1. in the same period. That is.9 (0.3 (0.

10 6 . 2007). rentier-state theorists do not examine the possibility that mineral abundance can be central to the development of manufacturing industry in particular. Norway.. geological knowledge and the technologies of mineral extraction. in earlier stages of development. If that is the case. the introduction of a dual-track growth strategy may be promising. Other authors explore how the development of natural resources led to increasingly high-tech industrial production in Sweden and Finland during the 19th and 20th centuries. especially to elites. Addressing the Challenges of Growth in Mineral-Abundant Countries: The Role of DualTrack Development Strategies Since the role of the government is generally more pronounced in oil and mineral-rich less developed countries. In general. By definition. the distribution of rents and privileges. by choosing oil booms as the point in which state formation takes place in latedeveloping oil economies. lessdiversified economies. more mineral-dominant. this strategy can be seen as a transitional path to more growth-enhancing institutional reforms. risk-taking and dynamic producers.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 Third. For instance. The basic idea of this strategy is to promote an emerging dynamic sector run (Track 1) where competition and risk-taking are promoted while maintaining the bulk of the protected and/or distorted sectors. In historical terms.11 The key policy question to ask is why natural-resource revenues are used in ways that sustain economic growth and diversification in some countries and not in others. then it makes sense to ask why. Karl’s model is subject to selection bias. Lack of economic diversification and poor economic growth are why economies are mineral dependent. the Netherlands. Australia and Malaysia were. Sweden. In this context. They demonstrate how large-scale investments in exploration. Such a dual-track strategy postpones confrontation with established rent-seekers while the dynamic sector drives competitive diversification of the economy and also builds a pro-reform political constituency. the US. However. 1999). 11 Blomström & Kokko (2007). suggesting a positive correlation between resource abundance and poor economic growth. Examples of Track 1 strategies are export-processing zones and industrial parks. for example. Wright & Czelusta (2007) examine how and why technological development and collective learning positively affected the development of natural resources in the US economy. For instance. 2008). often in protected agriculture and industrial sectors with aim of reducing social tensions and maintaining political stability (Track 2). Canada. The problem of selection bias renders many of the econometric studies. most countries that do not have a diversified agricultural and manufacturing base become mineral dependent. Much of the rent-seeking indeed leads to the creation of ineffective public spending and subsidisation. transportation. refining and utilisation in natural resources contributed to long-run economic growth and industrialisation in the US. there is likely to be significant amounts of pressure for patronage among contending groups and classes. political conflicts or past economic policies prevented growth in some mineral dependent economies and not in others. almost all countries began as mineral-dominant economies. Not only that. a trade-off between economic growth and political stability can emerge since those with access to state resources may be politically powerful but not necessarily effective. In such cases. natural resources generally played a growthenhancing role in stimulating capital accumulation and growth throughout the now advanced countries in the period 1870-1914 (Findlay & Lundhal.10 Finally. The main challenge of this strategy is to insulate/ring-fence the Track 1 sector from political and clientelist predation and capture. spurious (Brunnschweiler. is often central to the maintenance of political stability (North et al.

rentseeking and centralised interventionism and that these processes are necessarily productivity. one important threshold for this strategy to work in mineral-abundant economies would appear be the existence and/or construction of viable national political parties. Strong. Thus. Jonathan Di John School of Oriental and African Studies (SOAS). which reduces the possibilities that significant politically-destabilising horizontal inequalities will develop. On China.12 What is noteworthy in all these cases is the existence of strong national political parties and organisations underpinning executive authority. China and Mauritius. Moreover. they also provide a focal point around which collective action and lobbying can occur in a relatively predictable manner. A further exploration of which threshold effects are decisive in affecting the development path of resource-rich developing countries may provide some useful policy insights. 7 .and growth-restricting is not supported by comparative or historical evidence. On Indonesia. see Qian (2003). University of London 12 On Mauritius. Much more research is needed to examine why some economies are able to effectively use mineral and fuel rents in productive ways. Indonesia. On Malaysia. all of which are neglected in much of the research-curse literature. see Findlay & Wellisz (1993).Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 There are a range of countries that have attempted dual-track-strategies. Explaining governance and state capacity in such economies needs to be consistent with this basic evidence. see Flatters & Jenkins (1986). because national parties need to build crossethnic and cross-regional alliances. see Bruton (1992). The extent to which mineral and fuel abundance generate developmental outcomes depends largely on the nature of the state and politics as well as the structure of ownership in the export sector. These include Malaysia. Conclusion: The proposition that oil abundance induces extraordinary corruption. They also are central to providing the institutional mechanisms for distributing patronage to regional elites and to important political constituencies in ways that either prevent challenges to authority and/or maintain cohesion of the ruling coalition. Similar levels of state centralisation and corruption coincided with cycles of growth and stagnation in mineral and fuel-dependent economies. disciplined national parties not only enable the state to centralise patronage and make credible bargains and side-payments to contending groups.

D. Weingast (2007).. The Political Economy of Poverty. North. & K. A. United Nations University/World Institute of Development Economics Research. in G. Wallis. Natural Resources: Neither Curse nor Destiny. Journal of Development Studies. Washington DC. (1997). & S.). On the Brink: Weak States and US National Security. Bruton. & A. PA. Studies in the Economic History of the Middle East. Maloney (Eds. P. University of California Press. Blomström. Natural Resources and the Macro Economy. Palgrave. F. & M..). & S. D. 961-986. Economic Policy and Performance’. Engerman.. Di John. Institutions and Differential Paths to Growth among New World Economies: A View from Economic Historians in the United States’. (1970). R. Cambridge. Jomo (Eds. Oxford University Press.). Stanford University Press. p.) (2000). Equity and Growth: Sri Lanka and Malaysia. & Associates (1988). CGD Report 2004. nr 43 (6). S. R. in M. How Latin America Fell Behind: Essays on the Economic Histories of Brazil and Mexico. (2007). ‘Limited Access Orders in the Developing World: A New Approach to the Problems of Development’. Natural Resources: Neither Curse nor Destiny.). in D. ‘Oil Abundance and Violent Political Conflict: A Critical Assessment’.S. Strategic Factors in Economic Development. Cornell University. ‘Factor Endowments. Yale University Press. Institutions. Rent-Seeking and Economic Development. ’The Patterns and Problems of Economic Development in Rentier States’. (1930). Maloney (Eds. Brunnschweiler. Washington DC. in D.. Gelb. Findlay. ‘Resource-Led Growth: A Long-Term Perspective’. Sokoloff (1997). Khan. Rents. (1992). MIT Press. NY. H. Basingstoke. Jenkins (1986). Wellisz (1993). ‘Patterns of Rent-extraction and Deployment in Developing Countries: Implications for Governance. Mahdavy. ‘Cursing the Blessings? Natural Resource Abundance. Di John. N. Oil Windfalls: Blessing or Curse?. (2007. Lundhal (1999). Center for Global Development (2004). 1800-1914. S. From Windfall to Curse? Oil and Industrialization in Venezuela. University Park.. J. MA. Oxford University Press. Lederman & W. Cambridge. Oxford University Press. The World Bank. Harvard Institute for International Development. WIDER Working Paper 162. New York State School of Industrial and Labor Relations. H. & W. Penn State University Press. & K. Cook (Ed. Webb & B. Innis. The Political Economy of Poverty.). M. New Haven.. ‘Trade Structure and Growth’. Helsinki.. (2009). The Paradox of Plenty: Oil Booms and Petro States. Advancing Development: Core Themes in Global Economics. T. Cambridge University Press. Ithaca. Washington DC & Palo Alto. J. nr 36 (3). London. & G. Kaldor. in Stephen Haber (Ed.. Maloney (2007).. Cambridge. 1920 to the Present. M. The World Bank & Stanford University Press. Flatters. Lederman & W. Kokko (2007). Berkeley. Shorrocks (Eds. Mavrotas & A. J. Palo Alto: Stanford University Press. MA. World Development. Findlay. Stanford.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 Bibliographical References Auty.. Karl. Oxford University Press. ‘From Natural Resources to High-Tech Production: The Evolution of Industrial Competitiveness in Sweden and Finland’. Neary. New York. C. CGD. Lederman. Equity and Growth: Five Small Open Economies. van Wijnbergen (1986). H. World Bank 8 . The Fur Trade in Canada. Trade Policy in Indonesia. R. (2008). Oxford. (1967). and Economic Growth’.).

National Bureau of Economic Research. Cambridge. Washington DC & Palo Alto. Watkins. M. M. E. Maloney (Eds. (1990). Princeton. in Daniel Lederman & William F. 337-56. The World Bank. Washington DC. J. Ross. & J. ‘What Do We Know About Natural Resources and Civil War?’. (1963). The World Bank & Stanford University Press. Governing the Commons: The Evolution of Institutions for Collective Action. Warner (1995). & A. Canadian Journal of Economics and Political Science. (2004).). 9 . 29. ‘How Reform Worked in China’. Washington DC. ‘Resource-Based Growth: Past and Present’. ‘Extractive Sectors and the Poor’. Princeton University. G. Natural Resources: Neither Curse nor Destiny. Oxfam America Report. M. (2001). NBER Working Paper nr 5398. ‘Natural Resource Abundance and Economic Growth’. (2003). nr 41 (3). Y. p. Sachs.Area: Sub-Saharan Africa ARI 172/2010 Date: 15/12/2010 Policy Research Working Paper nr 4359. Ross. An Oxfam America Report. Czelusta (2007). Journal of Peace Research. NJ. Qian.). in D. In Search of Prosperity: Analytic Narratives of Economic Growth. vol.. ‘A Staple Theory of Growth’. 141-158.. Cambridge: Cambridge University Press. MA. nr 2. p. Ostrom. Wright. Rodrik (Ed.